This series of documents were developed by scientists in the University of Plymouth’s Deep Sea Conservation Research Unit as part of a pipeline to assist delegates of the ‘Habitat Conservation and Marine Spatial Planning’ workshop at the Deep Ocean Collective Solution Accelerator Meeting (Scripps Institution of Oceanography, 2-5th October 2023) in developing regional benthic habitat classifications. The pipeline mirrors that used in McQuaid et al. (2023).
The repository is designed to be downloaded as a zip folder and for the scripts to be run sequentially (steps 1 to 3) in the same R Studio session. This document is not designed to provide a thorough background in the theory of non-hierarchical habitat classification, nor the variable selection process. For this, we recommend reading Howell (2010) or the abovementioned McQuaid et al. (2023).
Should you notice any issues with the code or have any questions, please contact Dr Amelia Bridges at the University of Plymouth amelia.bridges@plymouth.ac.uk.
This guidance comprises 3 steps - accessing the data, clustering the input variables and combining the layers to create your benthic habitat classification.
To get started, follow these steps (NB: please make sure you have R and R Studio installed):
Unzip the folder in the location you want to work - e.g. Desktop/Documents etc.
Open the .html files and familiarize yourself with the process and what the outputs look like.
Double-click on the .Rproj file - this will open R Studio.
In the files panel in R Studio, open the ‘Step1_Accessing_the_data’ .qmd file*.
Work through the chunks of code in the script until you reach the end.
Without removing anything from your R environment or closing R, open the ‘Step2_Clustering_inputs’ .qmd file.
Work through the chunks of code in the script until you reach the end.
Open the ‘Step3_Combining_inputs’ .qmd file.
Work through the chunks of code in the script until you reach the end.
*The .qmd documents are the scripts used to create the .html files. These allow you to work through the process manually and alter variables/clusters/datasets if you wish.